Active reading for analysts

Analysts read more than they can remember. A single market question can pull in earnings calls, analyst reports, product docs, regulatory filings, and a dozen half-relevant articles. The constraint is rarely access to sources. It is holding onto what matters as you move through them.

Active reading is the discipline that addresses this directly. Instead of reading first and organizing later, you capture structured signal while you read, anchored to where each idea appeared. Understanding accumulates as you go, and the trail back to the evidence stays intact.

Passive reading loses provenance

Passive reading feels productive. You highlight, you nod, you move on. The cost shows up later: when you sit down to write, the highlights are detached from their context, and you cannot quickly say why a passage mattered or where it came from. Reconstructing that provenance is often slower than the original reading.

The deeper problem is that passive reading defers all the hard thinking to a single overloaded moment at the end. By then the source is cold.

A worked example

Take a passage like this, from an invented but typical filing: "The company recognized a $40M contract liability in Q3, tied primarily to the enterprise renewal signed in August, with the remainder to be recognized ratably through 2027." Skimmed and highlighted, that sentence tells you nothing when you come back to it three weeks later.

A capture built for reuse looks different. You still highlight the sentence, but you attach two things to it: a one-line note ("check whether the renewal terms match what the sales deck implied about enterprise pricing"), and, implicitly, why it mattered (a decision-relevant fact plus an open question, not just a number worth remembering). That is the whole unit: the passage, the note, the reason. It takes thirty seconds and it survives the trip back to your desk.

Capture signal, not text

Active reading is not highlighting more. It is capturing signal: the small interpretive act that turns a passage into something usable.

• A claim worth checking later ("management expects margin expansion in H2").
• A question the source raises but does not answer ("what assumption drives that estimate?").
• A connection to another source ("contradicts the supplier interview").
• A decision-relevant fact ("contract renewals concentrate in Q4").

Each of these is short, but each carries the interpretation that a bare highlight loses. Writing your own version of the point, rather than transcribing the passage verbatim, is also the part that does the remembering. A well-known study of note-taking during lectures found that people who paraphrased in their own words retained more of the material's structure than people who copied it down word for word (Mueller and Oppenheimer, 2014). The same mechanism holds mid-read: a paraphrased claim does more for you later than a faithful quote.

Keep capture in context

Every time you leave the source to record a thought, switching to a notes app, a doc, a chat window, you pay a context-switching tax and weaken the link between the idea and its origin. Over a long reading session those small breaks compound into fragmented attention and lost provenance.

This is not a new observation. The Cornell note-taking system, still taught in university study-skills programs, works on the same principle: capture happens next to the material as you take it in, not in a separate document assembled from memory afterward (Cornell University Learning Strategies Center). The practical rule for reading a source is the same one applied to reading a lecture: capture where you read, tag it with the kind of signal it is, and keep moving.

A repeatable reading pass

A reliable active-reading pass for a new source:

1. Frame first. Write the one question you need this source to inform. It filters everything else.
2. Read for signal. Capture claims, questions, connections, and decision-relevant facts in place, not every interesting sentence, only what advances the question.
3. Mark uncertainty. When a claim rests on an assumption or a number you cannot yet verify, flag it as something to check rather than accepting or discarding it.
4. Note relationships. When a source agrees with, extends, or contradicts another, record the link while both are fresh, and say why, not just that they connect. That habit, naming the reason for a link rather than leaving it implicit, is close to what the Zettelkasten note-taking method calls an explicit link (zettelkasten.de).
5. Close the loop. Before leaving, restate what this source changed about your view of the question.

What good active reading produces

Done well, active reading leaves you with more than notes. It leaves a set of source-grounded, interpreted signals that already point toward an output: claims ready to be checked, questions ready to be resolved, relationships ready to be synthesized.

None of this requires reading more slowly. The filing excerpt above took the same few seconds to read either way. What changes is whether those seconds leave behind a note you can use or a highlight you will have to re-derive later. Active reading is the version that leaves you with the note.